diff --git a/nomad/app/v1/routers/datasets.py b/nomad/app/v1/routers/datasets.py
index d65e6483cb93c112f70bcc5febd3685d70f4ddd5..7cc031d955a07153eb3b8b9d7392326f7c4d74e0 100644
--- a/nomad/app/v1/routers/datasets.py
+++ b/nomad/app/v1/routers/datasets.py
@@ -250,7 +250,7 @@ async def post_datasets(
if not empty:
processing.Calc._get_collection().update_many(
- mongo_query, {'$push': {'metadata.datasets': dataset.dataset_id}})
+ mongo_query, {'$push': {'datasets': dataset.dataset_id}})
update_by_query(
'''
if (ctx._source.datasets == null) {
@@ -309,7 +309,7 @@ async def delete_dataset(
if len(entry_ids) > 0:
processing.Calc._get_collection().update_many(
- mongo_query, {'$pull': {'metadata.datasets': dataset.dataset_id}})
+ mongo_query, {'$pull': {'datasets': dataset.dataset_id}})
update_by_query(
'''
int index = -1;
diff --git a/nomad/app/v1/routers/entries.py b/nomad/app/v1/routers/entries.py
index d846503f23b5e07536f79fefaba14a22d740a7f8..69834042b44b59e44a43c0b0428c1baf4dd71273 100644
--- a/nomad/app/v1/routers/entries.py
+++ b/nomad/app/v1/routers/entries.py
@@ -1327,7 +1327,7 @@ async def post_entry_metadata_edit(
# remove potentially empty old datasets
if removed_datasets is not None:
for dataset in removed_datasets:
- if proc.Calc.objects(metadata__datasets=dataset).first() is None:
+ if proc.Calc.objects(datasets=dataset).first() is None:
datamodel.Dataset.m_def.a_mongo.objects(dataset_id=dataset).delete()
return data
diff --git a/nomad/datamodel/datamodel.py b/nomad/datamodel/datamodel.py
index da80868ffb593f5210abb3f31c2b2e67e53d5bd2..52858ded2d88340bfd9f84de4e64ff60d71bffc5 100644
--- a/nomad/datamodel/datamodel.py
+++ b/nomad/datamodel/datamodel.py
@@ -574,7 +574,7 @@ class EntryMetadata(metainfo.MSection):
datasets = metainfo.Quantity(
type=dataset_reference, shape=['0..*'], default=[],
- categories=[MongoMetadata, EditableUserMetadata],
+ categories=[MongoEntryMetadata, EditableUserMetadata],
description='A list of user curated datasets this entry belongs to.',
a_elasticsearch=Elasticsearch(material_entry_type))
diff --git a/nomad/processing/data.py b/nomad/processing/data.py
index 89a2e6b5775fde7823aa8868a3c97fa215c37efe..4fccacbc35f99ac519765b7f59faf64f22d35cff 100644
--- a/nomad/processing/data.py
+++ b/nomad/processing/data.py
@@ -173,6 +173,7 @@ class Calc(Proc):
comment: a user provided comment for this entry
references: user provided references (URLs) for this entry
coauthors: a user provided list of co-authors
+ datasets: a list of user curated datasets this entry belongs to
metadata: the metadata record wit calc and user metadata, see :class:`EntryMetadata`
'''
@@ -193,6 +194,7 @@ class Calc(Proc):
references = ListField(StringField(), default=None)
coauthors = ListField(StringField(), default=None)
shared_with = ListField(StringField(), default=None)
+ datasets = ListField(StringField(), default=None)
metadata = DictField() # Stores user provided metadata and system metadata (not archive metadata)
@@ -207,7 +209,7 @@ class Calc(Proc):
('upload_id', 'nomad_version'),
'process_status',
'last_processing_time',
- 'metadata.datasets',
+ 'datasets',
'pid'
]
}
@@ -475,11 +477,9 @@ class Calc(Proc):
if self.upload is None:
logger.error('calculation upload does not exist')
- has_previous_metadata = bool(self.metadata)
-
# 1. Determine if we should parse or not
self.set_process_step('Determining action')
- if not self.upload.published or not has_previous_metadata:
+ if not self.upload.published or not self.nomad_version:
should_parse = True
else:
# This entry has already been published and has metadata.
@@ -1693,13 +1693,12 @@ class Upload(Proc):
# Handle datasets
dataset_ids: Set[str] = set()
for entry_dict in bundle_info['entries']:
- entry_metadata = entry_dict['metadata']
- entry_metadata_datasets = entry_metadata.get('datasets')
- if entry_metadata_datasets:
+ entry_datasets = entry_dict.get('datasets')
+ if entry_datasets:
if not include_datasets:
- entry_metadata['datasets'] = []
+ entry_dict['datasets'] = None
else:
- dataset_ids.update(entry_metadata_datasets)
+ dataset_ids.update(entry_datasets)
if include_datasets:
bundle_info['datasets'] = [
datamodel.Dataset.m_def.a_mongo.get(dataset_id=dataset_id).m_to_dict()
@@ -1854,7 +1853,6 @@ class Upload(Proc):
keys_exist(entry_dict, required_keys_entry_level, 'Missing key for entry: {key}')
assert entry_dict['process_status'] in ProcessStatus.STATUSES_NOT_PROCESSING, (
f'Invalid entry `process_status`')
- entry_metadata_dict = entry_dict['metadata']
# Check referential consistency
assert entry_dict['upload_id'] == self.upload_id, (
'Mismatching upload_id in entry definition')
@@ -1882,10 +1880,12 @@ class Upload(Proc):
# Instantiate an EntryMetadata object to validate the format
try:
if settings.include_datasets:
- entry.metadata['datasets'] = [
- dataset_id_mapping[id] for id in entry_metadata_dict.get('datasets', [])]
+ entry_datasets = entry_dict.get('datasets')
+ if entry_datasets:
+ entry.datasets = [
+ dataset_id_mapping[id] for id in entry_datasets] or None
else:
- entry.metadata['datasets'] = []
+ entry.datasets = None
entry.mongo_metadata(self)
# TODO: if we don't import archive files, should we still index something in ES?
except Exception as e:
diff --git a/tests/app/v1/routers/test_datasets.py b/tests/app/v1/routers/test_datasets.py
index 4630805eb9ddf1865d44ee3da216ef023236a7c3..efcf89f8ed9ee08ac3f301cc976e50964b91e152 100644
--- a/tests/app/v1/routers/test_datasets.py
+++ b/tests/app/v1/routers/test_datasets.py
@@ -152,7 +152,7 @@ def assert_dataset(dataset, query: Query = None, entries: List[str] = None, n_en
expected_n_entries = n_entries if dataset['dataset_type'] == 'owned' else 0
assert search_results.pagination.total == expected_n_entries
- assert processing.Calc.objects(metadata__datasets=dataset_id).count() == expected_n_entries
+ assert processing.Calc.objects(datasets=dataset_id).count() == expected_n_entries
def assert_dataset_deleted(dataset_id):
@@ -162,7 +162,7 @@ def assert_dataset_deleted(dataset_id):
search_results = search(
owner='admin', query={'datasets.dataset_id': dataset_id}, user_id=admin_user_id)
assert search_results.pagination.total == 0
- assert processing.Calc.objects(metadata__datasets=dataset_id).count() == 0
+ assert processing.Calc.objects(datasets=dataset_id).count() == 0
@pytest.mark.parametrize('query, size, status_code', [
diff --git a/tests/app/v1/routers/test_uploads.py b/tests/app/v1/routers/test_uploads.py
index 84c5900bd02fb94258f2f2469cdd53568518dcfb..85f41a7751b27cba5547a6012081f236fac82fb7 100644
--- a/tests/app/v1/routers/test_uploads.py
+++ b/tests/app/v1/routers/test_uploads.py
@@ -1262,7 +1262,7 @@ def test_post_upload_action_publish_to_central_nomad(
datamodel.Dataset(
dataset_id='dataset_id', dataset_name='dataset_name',
user_id=test_users_dict[user].user_id).a_mongo.save()
- calc.metadata['datasets'] = ['dataset_id']
+ calc.datasets = ['dataset_id']
calc.save()
# Finally, invoke the method to publish to central nomad
@@ -1288,7 +1288,7 @@ def test_post_upload_action_publish_to_central_nomad(
'upload_id', 'calc_id', 'upload_create_time', 'entry_create_time',
'last_processing_time', 'publish_time'):
assert new_calc_metadata_dict[k] == v, f'Metadata not matching: {k}'
- assert new_calc.metadata.get('datasets') == ['dataset_id']
+ assert new_calc.datasets == ['dataset_id']
assert old_upload.published_to[0] == config.oasis.central_nomad_deployment_id
assert new_upload.from_oasis and new_upload.oasis_deployment_id
diff --git a/tests/processing/test_data.py b/tests/processing/test_data.py
index 56f0fa1569979cf548001253eab57da47f358237..877de1e2aee1f85ad41c2fdb0f77668dba4099f5 100644
--- a/tests/processing/test_data.py
+++ b/tests/processing/test_data.py
@@ -278,7 +278,7 @@ def test_oasis_upload_processing(proc_infra, oasis_example_uploaded: Tuple[str,
assert_processing(upload, published=True)
calc = Calc.objects(upload_id='oasis_upload_id').first()
assert calc.calc_id == 'test_calc_id'
- assert calc.metadata['datasets'] == ['oasis_dataset_1', 'cn_dataset_2']
+ assert calc.datasets == ['oasis_dataset_1', 'cn_dataset_2']
@pytest.mark.timeout(config.tests.default_timeout)